Odaily Planet Daily reports that AI security startup Depthfirst claims its proprietary AI model outperforms Anthropic’s latest Mythos model in detecting code vulnerabilities, uncovering more critical security flaws at approximately one-tenth the cost, drawing attention from the cybersecurity industry.
According to the company, Mythos claimed to have discovered numerous critical vulnerabilities in core internet infrastructure code one month before its launch, while Depthfirst stated that its model, at a significantly lower cost (approximately $1,000 compared to $10,000), identified several additional high-severity vulnerabilities that Mythos missed.
Depthfirst CEO Qasim Mithani stated that the company has improved vulnerability detection efficiency using "single-task optimized AI models," significantly reducing security analysis costs while enhancing coverage depth.
The company completed an $80 million funding round in March, achieving a $580 million valuation, and simultaneously launched the Open Defense Initiative, providing $5 million in AI detection credits to open-source developers and critical infrastructure projects for vulnerability scanning and security audits. (Forbes)